Overview

Shannon solved the problem for one sender and one receiver. But what if multiple people are talking at once? (Like a cocktail party or a Wi-Fi network).

Core Idea

Interference: In a network, my signal is your noise. The goal is to manage this interference to maximize the total throughput.

Formal Definition (if applicable)

Multiple Access Channel (MAC): Many senders, one receiver (e.g., Cell phones talking to a tower). Broadcast Channel (BC): One sender, many receivers (e.g., TV tower, or a teacher talking to a class).

Intuition

  • Time Division (TDMA): Take turns speaking.
  • Frequency Division (FDMA): Speak at different pitches.
  • Code Division (CDMA): Speak different languages at the same time (the brain filters it out).

Examples

  • The Internet: A mesh of routers.
  • Sensor Networks: Thousands of tiny sensors relaying data.
  • MIMO (Multiple Input Multiple Output): Using multiple antennas to send multiple streams of data at once (Spatial Multiplexing).

Common Misconceptions

  • “Interference is always bad.” (Network Coding shows that sometimes mixing signals is good. If A wants to send to B, and B to A, they can send A+B to a relay, saving a slot).
  • Relay Channel: Using a helper to boost the signal.
  • Interference Alignment: Cleverly aligning signals so they cancel out at the unintended receiver.

Applications

  • 5G/6G: Designing the next generation of mobile networks.
  • Ad-hoc Networks: Military or disaster relief comms without infrastructure.

Criticism / Limitations

Most problems in network information theory are still open (unsolved). We don’t know the capacity of even simple 3-node networks.

Further Reading

  • El Gamal & Kim, Network Information Theory